At the end of my long post about getting a 1968 BMW R50/2 back on the road, the bike was running and riding, but not "fast". Every time I would work on it, it would start easy, and idle fine. If I set off riding, though, it would start to miss, stutter, lose power, and just generally run rough any time the throttle was opened too far.
I had to leave it for a couple weeks, but when I finally got back to it, I pulled the spark plugs. Good news: they haven't loosened on their own again in the several times I've pulled them to inspect since. I must have just not torqued the fresh crush washers correctly the first couple times. Bad news: the right cylinder was running very rich.
I started to wonder if the fuel filter might not have caught some silt from the old tank, so I pulled the carburetor apart to clean it. Half way into disassembly, I found the answer to the richness: the main jet had unscrewed from its seat. This is actually a common enough problem with these old Bing carburetors that I had read a few posts about it on some forums. I finished disassembly, cleaned everything, and reassembled, making sure that jet was well-torqued.
The test-ride afterward started much better than the last one. I had much less stuttering, much more power. But, as I tried to get into the higher revs, the roughness came back.
I've been skeptical of the timing advance unit. Was it not advancing as the RPMs increased? A friend brought over an ignition timing light, and we pointed it at the viewing window. We knew there was a problem even before we rolled on the throttle: the idle timing was late. It was firing near top-dead-center, instead of the 9° before TDC indicated by the "S" mark on the flywheel. When we increased the RPMs, the timing did advance, but it stopped advancing before we reached the "F" mark denoting full advance. Aha!
It took us quite some time to adjust the points to get them to open at the S mark instead of far behind it. But when we finally did, the timing light indicated that everything was now happy. The S mark sat in the window at idle, and when we revved the engine, we would bring the F mark into view. At least, when cold, standing in the driveway.
Sadly the next test ride was horrible. Worse than the ride right after fixing the main jet. Worse than the ride at the end of my last video. Just no power anywhere, and a terrible hesitation just off idle.
I tore back in and made one more change to timing. There are markings on the magneto rotor that is mounted behind the points plate. When the engine is at the S mark, the markings on the rotor are supposed to be at exactly 12 o'clock and 6 o'clock. Mine weren't. In fact, one of the reasons it took us a while to adjust the points timing was that the rotor's markings were so far off that I couldn't get the matching mark in the plate that holds the points to line up with them. Trying to do so made all of the other adjustments hit their limits before the points opened at the right time. I had a think and a tinker and finally found a combination that got the notch and the markings closer (though still not aligned), while still opening the points correctly.
The next test ride was better. Not perfect, but at least improved. That is, until later in the ride when it started stumbling again. At that point, I started having weird issues where pulling on one throttle cable would speed the engine up, but pulling on the other would cause it to stall. Then it switched which cable had which effect. Finally on the last leg home, one of the cables stuck with the throttle open to highway cruising speed.
I'm a very scientific engineer. I really like to know exactly what is wrong, and to understand exactly how whatever action I decide to take is going to fix it. I like to take small steps, testing one change at a time. The more I work on this bike, the less this mantra seems to work. This comes from two directions.
One is that so many of the adjustments are interrelated. Changing the points gap changes the ignition timing. Two different methods designed to change the ignition timing may also change the points gap. Changing the idle mixture changes the idle speed. But at a different idle speed, you may need a different mixture. Round and round.
Direction two is the more pernicious. If any of the interrelated parts is misbehaving, adjustments in other parts can't be trusted. You can't trust the carburetor balance if the throttle cables stick or stretch. You can't trust the idle adjustment or mixture if the spark isn't strong. Et cetera.
So, I finally gave up, and decided to tackle the N things I was worried about all in one shot.
First up: patching the exhaust. Until you get underneath the bike, it looks like there is one exhaust pipe for each cylinder. This is mostly true, but also just behind the transmission is a crossover pipe linking the two. Right at the junction with the crossover in each pipe, there were holes and inch long by an eighth of an inch or so wide. We had partially patched them with some asbestos wrap and band clamps, but that wasn't a perfect seal to start with, and it had deteriorated over time.
From what I've read, engines are tuned for a specific exhaust flow. Cut the pipes too short, and the backpressure is wrong. Separate a linked exhaust, and the scavenged air inertia from the other cylinder is lost. I doubt these were causing my runnability problems, but if I was about to put on new equipment, I wanted the airflow correct before tuning.
Many will shake their head at me for my chosen fix, but for the time being, I've patched the exhaust holes with Extreme Heat JB Weld (and a bit of metal mesh window screen for support across the gap). I haven't learned to properly weld yet, so since the spot for the application is mostly hidden, and these pipes are not anywhere near showroom state, I went with the option available. It works! I noticed a big change in tone while sitting next to the engine balancing the carburetors after all this.
Next up, new throttle cables. I'm fairly sure that what was on there were the originals. I lubricated them, and they slid okay, but a little gritty, and with some extra spinginess caused by the sheath decomposing near the ferrules. When I got to looking at old photos of the bike, I realized that if they were the original cables, they were also too long. The bike had come with "high" handlebars, but had since been converted to "low" bars. So, the cables had much tighter bends in them than they should, to account for the extra length. New cables slide much easier, and more crisply.
Finally, all the ignition parts. Literally every post I've read about similar issues, and every commenter on my earlier video, has suggested that I need a new coil and/or condenser. I had guessed at the coil myself, due to a similar problem I had years ago with my V-Strom. When that bike's stator (also a big coil of wire) went bad, the shop had a terrible time diagnosing the issue, because it didn't test bad when the bike was cold. It was only when I had it hot while commuting down the highway that it suddenly stopped producing electricity. So I bought and installed a new coil. Because I knew that "extra points and condenser" was a checkbox on many old travel checklists, I already had a set on hand, so I installed those as well while I was in there.
And just to finally end my worries about the advance unit, I also installed a fresh grease felt to keep it properly lubricated.
But this is where everything suddenly got interesting. Remember those markings on the magneto rotor I mentioned earlier? Well, I decided that if I was going to have access to it while installing the coil/condenser/points, that I should try to correct it. I ordered the proper tool, and pulled it off of its cone mounting on the end of the camshaft, then set the engine to the S mark and reinstalled the rotor with the markings oriented correctly. I installed the rest of the new parts over top of it, and set about getting the new points gap and timing reset.
After a while, it seemed like my alterations to timing were having completely random effects. I would try to advance and end up behind. I'd try to undo that, but end up ahead again. That's when I remembered to check the timing on both cylinders, and learned I had the dreaded differential timing. One side fired earlier than the other. The "random effects" I was seeing were from choosing an adjustment based on the observation of one cylinder, then evaluating the adjustment based on the observation of the other cylinder.
I had watched half a dozen YouTube videos about setting timing on these bikes, and every one had stopped at, "Make sure you don't have differential timing." Even the owner's manual says that if you have more than 2° of timing difference between cylinders, you need a new advance unit (which is what houses the points cam). I had over 5°.
I scoured the Vintage BMW forums, and finally found references to a post by the mechanic who has fixed more of these /2 machines that anyone, Craig "Vech" Vechorik. A little more digging found a reply he had made to someone else's thread where he described his technique for fixing differential timing by making pencil marks on the flywheel to learn exactly where to tap on the bolt holding the timing advance unit onto the magneto rotor. The idea is that this fixes an incorrect seating of the rotor onto its cone mount on the end of the camshaft.
I tried it. You can watch me try it in a video I just posted to YouTube demonstrating the technique: Correcting Differential Timing on a BMW /2 Motorcycle. It worked! Differential timing: solved. I was finally able to set the rest of the timing after that.
I plugged in the spark plugs, and gave the starting lever a kick. I swear the sparks have never been so blue. I screwed them into their holes, turned on the gas, and gave it a kick for real. It started as easy as it always has. Hooray! And then it revved easier than it ever has. Yippee! Out onto the road…
The first ride, I can't lie, didn't put my worries to bed. There was still some hesitation off idle. Getting into higher revvs was iffy. But, I knew the carburetors wouldn't be set right, or sync'd to each other at all. This was the warm up ride to prepare for that work.
After I had returned home and set up cooling fans, I pulled out my final new tool. Every manual I have suggests tuning and balancing the carburetors by occasionally pulling one of the spark plug caps off to listen to just one cylinder firing. Every forum notes that this is terrible for the health of the coil, because the energy for sparking the disconnected plug has "nowhere to go". The solution is to screw an extension onto the end of the spark plugs, so that the connection to the inside of the cap is exposed. This lets you stop that plug from sparking by using a screwdriver to ground out the wire to the engine case instead of removing the cap. Two 40¢ screws, a hacksaw, and the screw-off buttons that the plugs came with, and I was in business.
The ride after tuning was glorious. The bike hasn't run that smoothly since the one lovely day I had making a big loop around New Hampshire. It has power. It will do high revs. Highway speed is no trouble. There is a tiny hesitation just off idle, but it's honestly possible that's just my lack of experience with slide carbs making my wrist try to open it too quickly. I've decided to run some Seafoam through the system for a tank before retuning again. Update from another ride a few days later: hesitation is gone, three cheers to Seafoam and/or muscle memory adjustment!
The R50/2 seems happy again. I took it out to run errands this morning, as I have been intending to all summer. Coffee from a local roaster, yogurt from the grocery store, and a smile on my face.
Categories: Motorcycle
Post Copyright © 2024 Bryan Fink